Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
This document is published in:
Experimental Dermatology, Vol. 22, nº 3 (2013), pp. 195-201
DOI: http://dx.doi.org/10.1111/exd.12097
© 2013 John Wiley & Sons.
The regenerative potential of fibroblasts in a new diabetes-induceddelayed humanised wound healing model
Lucıa Martınez-Santamarıa1,2,3, Claudio J. Conti4, Sara Llames3,5, Eva Garcıa3,5, Luisa Retamosa2,3,
Almudena Holguın2,3, Nuria Illera2,3, Blanca Duarte3,6, Lino Camblor7, Jose M. Llaneza7, Jose L. Jorcano1,2,3,6,
Fernando Larcher3,6,Alvaro Meana3,5, Marıa J. Escamez1,2,3and Marcela Del Rıo1,2,3
1Department of Bioengineering, Carlos III University (UC3M), Madrid, Spain;2Regenerative Medicine Unit, Epithelial Biomedicine Division,
CIEMAT, Madrid, Spain;3Centre for Biomedical Research on Rare Diseases (CIBERER), ISCIII, Valencia, Spain;4Department of Molecular and
Cellular Medicine, College of Medicine, Texas A&M Health Science Center, College Station, TX, USA;5Tissue Engineering Unit, CentroComunitario de Sangre y Tejidos de Asturias (CCST), Oviedo, Spain;6Cutaneous Diseases Modeling Unit, Epithelial Biomedicine Division,
CIEMAT, Madrid, Spain;7Department of Angiology and Vascular Surgery, Hospital Universitario Central de Asturias, Oviedo, Spain
Correspondence:Dr. Marcela del Rio, Bioengineering Department, Carlos III University (UC3M). CIEMAT CIBERER. Avda. de la Universidad, 30.
28911 Leganes, Madrid, Spain, Tel.: +34 91 3466051, Fax: +34 91 3466484, e mail: [email protected]; Dr. Marı́a JoséEscámez Toledano,Regenerative Medicine Unit, Epithelial Biomedicine Division, CIEMAT CIBERER. Carlos III University (UC3M). Av. Complutense 22 Edificio
70A.0.14. 28040 Madrid, Spain. Tel.: +34 91 4962526, e mail: [email protected]
Abstract:Cutaneous diabetic wounds greatly affect the quality of
life of patients, causing a substantial economic impact on the
healthcare system. The limited clinical success of conventional
treatments is mainly attributed to the lack of knowledge of the
pathogenic mechanisms related to chronic ulceration. Therefore,
management of diabetic ulcers remains a challenging clinical issue.
Within this context, reliable animal models that recapitulate
situations of impaired wound healing have become essential. In this
study, we established a newin vivohumanised model of delayed
wound healing in a diabetic context that reproduces the main
features of the human disease. Diabetes was induced by multiple
low doses of streptozotocin in bioengineered human skin engrafted
immunodeficient mice. The significant delay in wound closure
exhibited in diabetic wounds was mainly attributed to alterations in
the granulation tissue formation and resolution, involving defects
in wound bed maturation, vascularisation, inflammatory response
and collagen deposition. In the new model, a cell based wound
therapy consisting of the application of plasma derived fibrin
dermal scaffolds containing fibroblasts consistently improved the
healing response by triggering granulation tissue maturation and
further providing a suitable matrix for migrating keratinocytes
during wound re epithelialisation. The present preclinical wound
healing model was able to shed light on the biological processes
responsible for the improvement achieved, and these findings can
be extended for designing new therapeutic approaches with clinical
relevance.
Key words:animal models – delayed wound healing – diabetic wounds –
fibroblasts – tissue engineering
IntroductionTissue repair is accomplished by the close coordination of the
processes of inflammation, re epithelialisation, granulation tissue
formation and dermal remodelling that overlap in space and time
(1 3). Chronic non healing wounds, such as diabetic ulcers, fail to
proceed through these sequential events (4), leading to diminished
quality of life, frequent hospitalisation and increased morbidity
and mortality (5,6). Moreover, this pathologic healing condition
causes significant socio economic consequences for patients, their
families and the healthcare system (7).
A peripheral nerve dysfunction induced by sustained hyper
glycaemia, with or without coexisting ischaemia, combined with
repeated minor trauma seem to be the prime causes of diabetic
skin ulcers (8 10). In conjunction with the diabetic neuropathy,
several altered cellular and molecular processes, mainly related to
granulation tissue formation and resolution, have been linked to
impaired diabetic cutaneous wound healing (11).
Due to the limited understanding of the mechanisms responsi
ble for poor healing, conventional therapies are unable to guaran
tee an adequate and sustained regeneration of the damaged tissue
(12). More recently, fibroblasts administered as a part of bioengi
neered dermal or dermo epidermal substitutes offer new therapeu
tic possibilities in both animal models and clinical studies
(13 15). In particular, fibroblasts are able to grow and provide
multiple growth factors and extracellular matrix proteins (16 18)
that might be altered in diabetic wound beds. Consequently,
tissue engineering approaches are emerging as smart strategies for
the treatment of chronic cutaneous wounds such as diabetic
ulcers (19).
The study of diabetic wound healing in patients is limited by
obvious ethical considerations and the heterogeneity of the
disease. Thus, apart from a few studies performed on volunteers
(20 22), current knowledge of wound healing mainly stems from
the use of a broad variety of transgenic and mutant murine
models (23 26). Despite the unquestionable value of murine
models, innovative approaches must been explored in a human
ised context to obtain results of clinical relevance. To this end,
chimeric systems composed of skin of human origin, vascularised
by murine vessels, have been exploited by several groups, includ
ing ours (27 29). The skin humanised mouse model developed
by our laboratory is based on the permanent engraftment of a
fibrin based bioengineered human skin onto the back of
1
immunodeficient mice (29,30). The orthotopic bioengineered
skin transplantation methodology enables the generation of
numerous mice engrafted with a significant homogeneous area of
single donor derived stable human skin (31,32) and offers
multiple possibilities to faithfully recreate diverse cutaneous
pathological (33 35) and physiological processes (36,37).
In the present work, we report a newin vivomodel of delayed
wound healing in a diabetic context using the skin humanised
mouse model that resembles a clinically meaningful skin repair
deficient condition. Furthermore, the model has been used as a
preclinical platform to evaluate the effectiveness of fibroblast
containing fibrin based dermal matrices which considerably
enhanced tissue repair.
MethodsPrimary culture of human keratinocytes and fibroblastsCells from skin biopsies of three healthy donors were isolated by
mechanical and enzymatic digestion as previously described
(30,38,39). Donors were subjected to standard serological tests
according to national regulations (RD 1301/2006). Donors gave
their written informed consent for biopsies, and all experimental
procedures were conducted in accordance with the World Medical
Association Declaration of Helsinki and subsequent revisions.
Grafting of bioengineered human skin equivalents: the skinhumanised miceSix week old female immunodeficient nude mice (Rj: NMRI
Foxn1nu; Elevage Janvier, Le Genest Saint Isle, France) were
orthotopically grafted with bioengineered cutaneous equivalents.
This setting contains human keratinocytes (epidermal component)
seeded on a plasma derived fibrin matrix populated with live
human fibroblasts (dermal component). Grafting was performed
as previously described (29,30) under sterile conditions at the
CIEMAT Laboratory Animals Facility (European registration num
ber ES280790000183; Spanish registration number 28079 21 A).
All handling was carried out according to European and Spanish
laws and regulations on the protection and use of animals in
scientific research. Experimental procedures were approved by the
Animal Experimentation Ethical Committee of CIEMAT.
Wound healing experimental design in the diabetic skinhumanised mouse modelDiabetes was induced in skin humanised mice at 10 weeks post
grafting by intraperitoneal injections of streptozotocin (STZ;
Sigma Aldrich, St. Luis, MO, USA), dissolved in sodium citrate
buffer (0.01M; pH 4.5). Blood glucose levels were routinely mea
sured under non fasted conditions by tail vein sampling using an
Accu Check Blood Glucose Monitor (Roche Diagnostic, Indianap
olis, IN, USA). Mice whose blood glucose levels exceeded 250 mg/
dl on two consecutive measurements separated by an interval of
48 h were considered diabetic. Normoglycaemic mice were used as
controls.
Sustained release bovine insulin pellets (Linbit, LinShin, North
Scarborough, ON, Canada) were provided to severely diabetic
mice (blood glucose levels higher than 500 mg/dl) to make hyper
glycaemia compatible with life. Further information is provided in
Fig. S1.
Full thickness 2 mm excisional wounds were created in the sta
ble human skin graft using a biopsy punch (Pfm Medical Ag.,
Cologne, Germany). Circular excisioned tissue was harvested and
used as a reference for anatomopathological analysis. Clinical
follow up of the healing process was performed by using a camera
coupled to a stereomicroscope (Olympus, Barcelona, Spain).
Anatomopathological analysis was performed at three (n=4
mice per experimental group), seven (n=10) and 14 days
(n=3) postwounding. Sample harvesting and processing were
carried out as previously described (36,40). Serial 4lm cross sec
tions were obtained. The whole sample was sectioned to determine
the centre of the wound and adequately monitor the healing pro
cess. Haematoxylin and eosin staining (Thermo Shandon GmbH,
Darmstadt, Germany) was routinely performed in tissue sections.
Wound healing analysisThe re epithelialisation percentage was calculated by the formula:
1009 [(wound diameter epidermal gap)/wound diameter] (41).
Immunohistochemical and immunofluorescence procedures were
performed in tissue sections adjacent to the centre following stan
dard protocols.
To determine the human origin of epidermal and dermal cells,
species specific antibodies against involucrin (dilution 1:100; clone
SY5 mAb, Sigma, St Louis, MO, USA) and vimentin (dilution
1:50; clone V9 mAb, BioGenex, San Ramon, CA, USA) were used,
respectively.
Mature blood vessel density was determined using an antibody
specific forasmooth muscle actin (SMA; dilution 1:400; clone
1A4 mAb; Sigma).
Polymorphonuclear neutrophil (PMN) infiltration in the
wound bed was determined by labelling myeloperoxidase activity
(MPO; dilution 1:50; Hycult Biotechnology, Uden, the Nether
lands).
The nuclear antigen Ki67 (Thermo Scientific, Fremont, CA,
USA) was selected as a cell proliferation marker. In addition, an
antibody against keratin K10 (dilution 1:1000; Covance, Emeryville,
CA, USA) was used as an early differentiation marker of keratino
cytes. For Ki67 and keratin K10 staining, a heat induced epitope
retrieval treatment using 10 mMcitrate buffer was required.
Specific biotinylated secondary antibodies were purchased from
Jackson ImmunoResearch Laboratories (West Grove, PA, USA).
After colour development (Avidin Biotin Complex Vectastain Elite
kit; Vector Laboratories, Inc., Burlingame, CA, USA), sections
were counterstained with haematoxylin.
Nerve fibres were visualised by immunofluorescence using an
antibody against the pan neuronal marker protein gene product
9.5 (PGP 9.5; dilution 1:500; ABD Serotec, Oxford, UK) and a
FITC conjugated secondary antibody (Jackson Immuno Research
Laboratories). Cell nuclei were stained with DAPI (Merck,
Darmstadt, Germany). Microphotographs were taken using a flu
orescence microscope (Axioplan 2; Carl Zeiss GmbH, Jena, Ger
many).
Collagen deposition was assessed by picrosirius red staining
following standard procedures (42), and stained sections were
analysed by polarised light microscopy (U ANT filter; Olympus).
Treatment of diabetic wounds with bioengineered dermalequivalents in the skin humanised mouse modelFibrin based bioengineered dermis containing 106 live human
fibroblasts were applied covering diabetic wounds (n=4) and
were kept in place using TegadermTM
2
(3M Health Care, ST Paul,
MN, USA), as previously described (40). After 7 days, wound
samples were collected and analysed. An identical number of
diabetic wounds were treated with acellular fibrin gels.
Statistical analysisWound healing parameters were analysed in the different experi
mental groups, and means were compared using the two tailed
unpaired Student’sttest. Differences were considered statistically
significant whenP
The inflammatory response in the stroma of 7 day diabetic
wounds was associated with a massive infiltration of myeloperoxi
dase positive neutrophils flanking the fibrin plug (Fig. 3c). In con
trast, much fewer neutrophils were found in the granulation tissue
of control wounds, revealing a practically resolved inflammatory
phase.
The Ki67 proliferative index of basal keratinocytes was assessed in
unwounded and wounded regenerated human skin. Surprisingly, on
day 7, a comparable postwounding proliferative burst was observed in
both experimental groups (Fig. 3d,e). However, while well developed
migratory tongues, with no stratum corneum and decreasing expres
sion of K10, were observed in controls, diabetic wounds exhibited a
(a)
Control
Diabetic
**
100
300
500
Glucose % re epithelialisation
G ucose (mg/d )
10
30
50
70 % re-ep the a sat on
**
(d)
Contro
D abet c
Contro
(b)
D abet c
(c)
Figure 2.Impaired granulation tissue formation and dermal remodellingcontributed to delayed wound healing in the diabetic skin humanised mousemodel. (a) Percentage of re epithelialisation calculated by the formula: 1009[(wound diameter epidermal gap)/wound diameter]. Values are expressed asmean SD (n=10 mice in each group).**P
fully keratinised acanthotic epithelium with a concomitant expression
of keratin K10 reaching the edge of the wound (Fig. 3e, f), evidencing
impaired migratory activity of keratinocytes.
Noticeably, long term analysis revealed that 7 day postwoun
ding defects were still evident, albeit attenuated, in completely
re epithelialised 14 day diabetic wounds (Fig. S2C).
Bioengineered dermis accelerated wound healing in diabeticwoundsDiabetic wounds in the skin humanised mouse model displayed
delayed granulation tissue formation with poor vascular and cellu
lar density, mimicking the pathological conditions that predispose
diabetic patients to develop cutaneous ulcers (4,44,45). With the
aim of correcting these biological alterations, the therapeutic
potential of tissue engineered dermal equivalents containing
human fibroblasts assembled in a three dimensional fibrin matrix
was evaluated in a diabetic context (Fig. S4). In the preclinical
model, a significant healing improvement was seen in diabetic
wounds at 7 days post treatment (Fig. 4a,b), achieving a
re epithelialisation rate comparable to that found in untreated
control wounds (Fig. 2a). Moreover, a granulation tissue rich in
collagen was observed in diabetic wounds after the treatment with
bioengineered dermal equivalents (Fig. 4c). Additionally, fibro
blast containing bioengineered dermis seemed to exert an angio
genic effect during granulation tissue formation (Fig. 4d,e), and
indeed, vascular density in treated diabetic wounds was compara
ble to that found in untreated controls (Fig. 3a). Furthermore, no
noticeable effects on granulation tissue formation were detected
when fibroblasts were not included in the fibrin gels used for
wound treatment (Fig. 4c e).
Our results demonstrated, in a preclinical model, the effective
ness of bioengineered dermal equivalents in diabetic wounds,
particularly through the promotion of wound bed maturation.
DiscussionDiabetes affects approximately 2% of the world population in
developed countries and its incidence increases in relation to
current lifestyle (46). Diabetic patients frequently suffer from
impaired wound healing and are consequently prone to develop
cutaneous ulcers with a high rate of recurrence that often become
chronic (5,6,47). The development of effective treatments for heal
ing chronic ulcers largely depends on understanding the patho
genic mechanisms involved. This poses the need for reliable
animal models that recapitulate situations of impaired wound
healing (23,48). The db/db mouse model presents delayed cellular
infiltration, granulation tissue formation and collagen deposition,
as well as prolonged persistence of neutrophils and reduced angio
genesis (24,26). These defects have also been depicted in our
model. However, extrapolation of results coming from rodent and
other small mammal models should be interpreted with caution,
due to major architectural and functional differences with human
epidermis (49). In fact, wound healing in mice is a process that
occurs mainly at the expense of contraction, a mechanism of
minor relevance in human skin (40,50). To overcome these
problems, several groups, including ours, have developed a skin
humanised mouse model (28 30,51). Specifically, the model devel
oped in our laboratory faithfully reproduces the main features of
human healing at both structural and functional levels (36,40).
Our preclinical humanised wound healing system comprises two
processes of skin regeneration. The first entails the grafting of bio
engineered human skin onto 12 mm excisional wounds in immu
nodeficient mice (29,30). The second regeneration process
involves the closure of 2 mm small full thickness wounds per
formed in mature, quiescent and stably engrafted human skin
(10 12 weeks after grafting). Closure of these small wounds is
carried out by human keratinocytes and human fibroblasts with
minimal contraction contributing to wound resolution (40).
To establish a model of wound healing impairment in the skin
humanised mouse model, a pharmacological induced diabetes
protocol based on STZ administration was employed. Both sur
vival and hyperglycaemia were modulated by a subcutaneous insu
lin implant that guaranteed elevated but life compatible levels of
(b)
20
40
60
80
100
**
*
% re-ep the a sat on
(e)
5
10
15
20
25
30
SMA pos t ve
vesse s/f e d
**
Bioengineered dermis treatedUntreated Fibrin gel treated
Untreated
(d)(c)(a)
Fibrin gel-treated
Bioengineered dermis-treated
Figure 4.Fibroblast containing bioengineered dermis enhanced wound healing inthe diabetic skin humanised mouse model. (a) Representative photographs (scalebar=1 mm) and composite pictures showing panoramic view of 7 day diabeticwounds stained with H&E (scale bar=500lm). Accelerated wound closure wasevidenced in diabetic wounds after treatment by a reduction of the epidermalgap distance (double arrowhead). Dashed black lines delimit the wound margins.(b) Percentage of re epithelialisation in diabetic wounds. Results are expressed asmean SD (n=4 mice in each group).**P
blood glucose during a period of time adequate to cause an
impact on the skin.
Hyperglycaemia induced nerve damage resulting in peripheral
neuropathy is a common complication in diabetes (10). As a
result of metabolic abnormalities, these patients may also present
endothelial dysfunction and alterations in the vascular supply to
organs, including the skin (8,52). Both decreased innervation and
vascularisation were exhibited on the regenerated human skin of
diabetic mice exposed to hyperglycaemia for 6 weeks, proving the
suitability of our protocol. Moreover, the significantly delayed
wound closure in STZ treated mice reproduces the healing
impairment associated with diabetes in patients (4,11) and in
other diabetic animal models (24,26).
Defects associated with granulation tissue formation and their
subsequent resolution seem to be responsible for the delayed heal
ing process in the present model. Specifically, poorly cellularised
7 day diabetic wound beds with reduced collagen deposition
resulted in an inadequate matrix for cell migration. These altera
tions correlate with the defective fibrinolysis and loose extracellu
lar matrix deposition described in diabetic patients (44,53). It has
been proposed that glycation of proteins, such as fibrinogen or
enzymes related to fibrin plug degradation, delays its resolution
and thus adversely affects migration of epithelial tongues in dia
betic wounds (54). In addition, alterations of the subepidermal
vascular plexus of diabetic mice observed prior to wound creation
were further evidenced during the angiogenic response triggered
by the healing stimulus. Thus, inadequately vascularised granula
tion tissue might not provide the metabolic support required for
cell recruitment, hampering the tissue repair process as described
in patients (45) and in other diabetic animal models (26).
In the new diabetic model herein described, a prolonged inflam
matory response, mainly consisting of neutrophils, could be gener
ating a proteolytic and pro oxidant environment in the wound bed,
thus contributing to the delayed healing progress (55). Indeed, a
sustained presence of polymorphonuclear granulocytes has been
reported in patients and animal models of diabetes (56,57).
Controversial data concerning keratinocyte proliferation and
hyperglycaemia have been reported. Although prolonged exposure
to high glucose inhibitsin vitro
6
human keratinocyte proliferation
(58,59), other evidences support that keratinocytes at the chronic
ulcer edge are hyperproliferative (60). In the model described here, a
postinjury keratinocyte proliferation burst without significant differ
ences associated with diabetes was observed. We thus hypothesise
that epithelial cells of the diabetic wound edges were capable of
responding to proliferative signals but epidermal migration might be
compromised because of the immaturity of the wound bed. As a
result, keratinocytes still expressing markers of differentiation accu
mulated at the edge of the wound, forming an acanthotic epithelium.
Long term analysis revealed that defects responsible for delayed
healing were attenuated in completely re epithelialised 14 day
wounds, thus wound chronification was not attained in the dia
betic skin humanised mouse model. Limiting factors, such as, the
acute toxicity exerted by STZ and the use of cells from healthy
donors employed for assembling the bioengineered human skin
equivalents, could explain the lack of chronicity.
As diabetic ulcers respond poorly to conventional treatments
(12), new approaches are needed. Advanced cell therapies using
healthy allogenic fibroblasts emerge as attractive strategies to over
come the senescence of fibroblasts present in chronic wound beds
(61,62). Intracellular signalling activated via cytokines, growth fac
tors and proteolytic enzymes secreted by allogenic healthy fibro
blasts seems to stimulate the recipient’s own wound bed derived
skin cells and therefore crucial processes frequently altered in
diabetic wounds, such as re epithelialisation and angiogenesis, are
activated (16 18). In addition, human fibroblasts are an easy to
handle, accessible cell source with acceptable cell yields obtained
from a relatively small skin biopsy. These properties, together with
their low immunogenicity (63,64) make fibroblasts a suitable tar
get in tissue engineering for the treatment of chronic wounds of
different aetiology (65).
Furthermore, the use of fibrin, unlike other biomaterials such
as collagen, preserves the viability and functionality of epidermal
stem cells (29,30,32,66,67), because it is a reservoir of cytokines,
clotting, cell adhesion and growth factors. Plasma derived fibrin,
unlike commercial purified fibrinogen preparations, contains addi
tional factors such as fibronectin or thrombospondin that may
contribute to keratinocyte adherence and survival (39,68). More
over, fibrin provides a suitable three dimensional scaffold to pro
mote migration, proliferation and differentiation of the cells in
the wound bed (69).
In the present work, the skin humanised mouse model has been
exploited as an ideal preclinical platform to evaluate the effective
ness of fibroblast containing fibrin based dermal scaffolds on
wound healing in a diabetic context. After 7 days of treatment, a
noteworthy enhancement of wound closure was achieved in dia
betic wounds, showing re epithelialisation rates comparable to
untreated control wounds. Interestingly, treatment with bioengi
neered dermis improved diabetic wounds by promoting wound
bed maturation. As these improvements were not achieved when
fibroblasts were not included in the fibrin matrix and these cells
are a well known source of soluble mediators, we suggest
fibroblasts induced maturation of diabetic wounds in terms of
collagen deposition and wound bed vascularisation. Consequently,
the granulation tissue of improved quality would provide a suit
able matrix for keratinocyte migration, explaining the differences
observed in wound closure after treatment.
Clinical benefits of tissue engineered products have been exten
sively demonstrated (19), although their use is considerably lim
ited due to the scarce availability and high cost associated with its
manufacture and development (70). In this context, the self made,
non commercial, bioengineered dermal setting we report herein
appears to be an attractive therapeutic alternative for diabetic
wounds not only in terms of efficacy, but also because of the sub
stantial cost savings.
In conclusion, the proposed model of delayed wound healing
emerges as a suitable preclinical tool to evaluate clinically mean
ingful innovative therapeutic approaches in the field of dermatol
ogy and also to provide a better understanding of the biological
mechanisms involved in wound healing improvement after a treat
ment. Furthermore, we have demonstrated that bioengineered
fibroblast containing dermal substitutes are a powerful and trust
worthy tool for the treatment of diabetic wounds in a preclinical
context. The bioengineered dermis may represent a realistic, effi
cient and inexpensive strategy to overcome the granulation tissue
defects frequently associated not only with diabetic wounds but
also with ulcers of different aetiology.
AcknowledgementsLMS and MJE performed the experiments, and together with MDR and
CJC designed the research study, analysed the data and wrote the
manuscript. SL, EG, LC, JML and AM manufactured the bioengineered
skin equivalents. LR, AH, NI and BD performed mice grafting and
tissue processing. FL and JLJ contributed to data interpretation and
experimental design. All the authors read and approved the final manu-
script. We especially thank our technicians I. Santos and F. Sanchez for
histology assistance, and J. Martınez and E. Almeida for animal
care. This work was supported by grants from the Science and Innova-
tion Ministry of Spain (SAF2010-16976), from the European VI Frame-
work Programme (LSHB-CT-512102), from Comunidad de Madrid
(S2010/BMD-2420; CELLCAM) and from Fundacion Ramon Areces
(CIVP16A1864).
Conflict of interestsThe authors have no conflicting financial interests.
References1 Martin P. Science 1997:276:75–81.2 Singer A J, Clark R A. N Engl J Med 1999:341:738–746.
3 Braiman Wiksman L, Solomonik I, Spira Ret al.Toxicol Pathol 2007:35: 767–779.
4 Blakytny R, Jude E. Diabet Med 2006: 23:594–608.
5 Carrington A L, Abbott C A, Griffiths Jet al.Diabetes Care 2001:24: 216–221.
6 Alvarsson A, Sandgren B, Wendel Cet al.Cardiovasc Diabetol 2012:11: 18.
7 Boulton A J, Vileikyte L, Ragnarson Tennvall Get al.Lancet 2005:366: 1719–1724.
8 Chao C Y, Cheing G L. Diabetes Metab Res Rev2009:25: 604–614.
9 Behm B, Schreml S, Landthaler Met al.J EurAcad Dermatol Venereol 2012:26: 1203–1211.
10 Callaghan B C, Cheng H T, Stables C Let al.Lancet Neurol 2012:11: 521–534.
11 Falanga V. Lancet 2005:366: 1736–1743.12 O’Loughlin A, McIntosh C, Dinneen Set al.Int
J Low Extrem Wounds 2010:9:90–102.13 Truong A T, Kowal Vern A, Latenser B Aet al.
J Burns Wounds 2005:4: e4.14 Llames S, Garcia E, Garcia Vet al.Cell Tissue
Bank 2006:7:47–53.15 Clark R A, Ghosh K, Tonnesen M G. J Invest
Dermatol 2007:127: 1018–1029.16 Mansbridge J, Liu K, Patch R et al.Tissue Eng
1998:4: 403–414.17 Mansbridge J N, Liu K, Pinney R Eet al.Diabetes
Obes Metab 1999:1: 265–279.18 Pinney E, Liu K, Sheeman Bet al.J Cell Physiol
2000:183:74–82.19 Wong V W, Gurtner G C. Exp Dermatol 2012:
21: 729–734.20 Ashcroft G S, Greenwell Wild T, Horan M A
et al.Am J Pathol 1999:155: 1137–1146.21 Brem H, Stojadinovic O, Diegelmann R Fet al.
Mol Med 2007:13:30–39.22 Ud Din S, Perry D, Giddings Pet al.Exp Derma
tol 2012:21: 758–764.23 Davidson J. Wounds 2001:13:9–23.24 Wall S J, Bevan D, Thomas D W et al.J Invest
Dermatol 2002:119:91–98.25 Grose R, Werner S. Mol Biotechnol 2004:28:
147–166.26 Tkalcevic V I, Cuzic S, Parnham M Jet al.Toxi
col Pathol 2009:37: 183–192.27 Demarchez M, Sengel P, Prunieras M. Dev Biol
1986:113:90–96.
28 Pouliot R, Larouche D, Auger F Aet al.Transplantation 2002:73: 1751–1757.
29 Del Rio M, Larcher F, Serrano Fet al.Hum GeneTher 2002:13: 959–968.
30 Llames S G, Del Rio M, Larcher Fet al.Transplantation 2004:77: 350–355.
31 Garcia M, Escamez M J, Carretero Met al.MolCarcinog 2007:46: 741–745.
32 Larcher F, Dellambra E, Rico Let al.Mol Ther2007:15: 1670–1676.
33 Guerrero Aspizua S, Garcia M, Murillas Ret al.Am J Pathol 2010:177: 3112–3124.
34 Garcia M, Larcher F, Hickerson R Pet al.J InvestDermatol 2011:131: 1053–1060.
35 Aufenvenne K, Rice R H, Hausser Iet al.J InvestDermatol 2012:132: 1918–1921.
36 Escamez M J, Garcia M, Larcher Fet al.J InvestDermatol 2004:123: 1182–1191.
37 Garcia M, Llames S, Garcia Eet al.Am J Pathol2010:177: 865–872.
38 Rheinwald J G, Green H. Cell 1975:6: 331–343.39 Meana A, Iglesias J, Del Rio M etal.Burns
1998:24: 621–630.40 Escamez M J, Carretero M, Garcia Met al.J
Invest Dermatol 2008:128: 1565–1575.41 Staiano Coico L, Krueger J G, Rubin J Set al.J
Exp Med 1993:178: 865–878.42 Junqueira L, Bignolas G, Brentani R. Histochem J
1979:11: 447–455.43 Szkudelski T. Physiol Res 2001:50: 537–546.44 Sobel B E, Schneider D J. Cardiol Clin 2004:22:
511–526.45 Cui T, Kirsner R, Li J. Angiogenesis in chronic
wounds. In: Sen C, ed. Advances in Wound Care:Volume 1 Translational Medicine: From Benchtopto Bedside to Community and Back. New Rochelle,NY: Mary Ann Liebert, Inc., 2010: 347–352.
46 Wild S, Roglic G, Green A et al.Diabetes Care2004:27: 1047–1053.
47 Brem H, Tomic Canic M. J Clin Invest 2007:117: 1219–1222.
48 Rees D A, Alcolado J C. Diabet Med 2005:22:359–370.
49 Sullivan T P, Eaglstein W H, Davis S Cet al.Wound Repair Regen 2001:9:66–76.
50 Wong V, Sorkin M, Glotzbach J et al.J BiomedBiotechnol 2011:2011: 969618.
51 Geer D J, Swartz D D, Andreadis S T. Tissue Eng2004:10: 1006–1017.
52 Dinh T, Veves A. Curr Pharm Des 2005:11:2301–2309.
53 DunnE J, Philippou H, Ariens R Aet al.Diabetologia 2006:49: 1071–1080.
54 Peppa M, Stavroulakis P, Raptis S A. WoundRepair Regen 2009:17: 461–472.
55 Pierce G F. Am J Pathol 2001:159: 399–403.56 Loots M A, Lamme E N, Zeegelaar Jet al.J
Invest Dermatol 1998:111: 850–857.57 Wetzler C, Kampfer H, Stallmeyer B et al.J
Invest Dermatol 2000:115: 245–253.58 Spravchikov N, Sizyakov G, Gartsbein Met al.
Diabetes 2001:50: 1627–1635.59 Terashi H, Izumi K, Deveci Met al.Int Wound J
2005:2: 298–304.60 Usui M L, Mansbridge J N, Carter W Get al.J
Histochem Cytochem 2008:56: 687–696.61 Telgenhoff D, Shroot B. Cell Death Differ 2005:
12: 695–698.62 Wall I, Moseley R, Baird D et al.J Invest Derma
tol 2008:128: 2526–2540.63 Haniffa M A, Wang X N, Holtick U et al.J
Immunol 2007:179: 1595–1604.64 Sorrell J M, Caplan A I. Int Rev Cell Mol Biol
2009:276: 161–214.65 Wong T, McGrath J A, Navsaria H. Br J Dermatol
2007:156:1149–1155.66 Pellegrini G, Ranno R, Stracuzzi Get al.Trans
plantation 1999:68: 868–879.67 Ronfard V, Rives J M, Neveux Yet al.Transplan
tation 2000:70: 1588–1598.68 Clark R A. Thromb Haemost 2003:90: 1003–
1006.69 Geer D J, Swartz D D, Andreadis S T. Tissue Eng
2002:8: 787–798.70 Langer A, Rogowski W. BMC Health Serv Res
2009:9: 115.
Supporting Information
Figure S1.Establishment of a diabetic skin-human-ised mouse model.Figure S2.Delayed wound healing in the diabeticskin-humanised mouse model.Figure S3.Species-specific antibodies confirmed thehuman origin of keratinocytes and fibroblasts in theregenerated human skin and in wounds in theskin-humanised mouse model.Figure S4.
7
Treatment of diabetic wounds with bioen-gineered dermal equivalents in the skin-humanisedmouse model.